The problem is there seems to be research to support every view out there.
For example this study:
UK Biobank data
Gong and colleagues analyzed the data of 273,240 women and 228,957 men without dementia who were enrolled in the U.K. Biobank — a prospective population-based cohort — between 2006 and 2010. They assessed participants’ follow-up data from baseline until incidence of all-cause dementia, death or until the cutoff on Nov. 30, 2020.
Found that:
There was no significant difference in dementia risk between women who used HRT or not (aHR = 0.99; 95% CI, 0.9-1.09), or for HRT duration (aHR = 1; 95% CI, 0.98-1.01). However, there was a small reduction in dementia risk with older age at HRT initiation (aHR = 0.96; 95% CI, 0.95-0.98).
The things they found that did make a difference were social deprivation, age of onset of periods (younger than 12 or over 14), having a baby before 21, having a hysterectomy especially if after oophorectomy.
Pregnancy and abortion both reduced the rate as did having a reproductive span of 36 years or longer.